Enriching SO₄²⁻ Immobilization on α-Fe₂O₃ via Spatial Confinement for Robust NH₃-SCR Denitration

Abstract

The application of iron oxide to NH₃-SCR is attractive but largely hindered by its poor acid properties, and surface sulfation is proven to be a prominent way of enhancing the acidity. As such, the method of enriching the sulfate species on iron oxide is crucial for improving the NH₃-SCR performance. In the present study, by employing ammonium bisulfate (ABS) as the source of gaseous SO₂ for the purpose of trapping, we reported an effective strategy for enhancing the SO₄²⁻ immobilization on α-Fe₂O₃ catalyst via spatial confinement in a mesoporous SBA-15 framework. Interestingly, although the presence of the mesopore channel had an adverse effect on the ABS decomposition, which was expected to produce less available SO₂, the measured SO₄²⁻ immobilized on α-Fe₂O₃ in the mesoporous SBA-15 system was significantly greater than that of the regular SiO₂, demonstrating the promoting effect of the spatial confinement on the SO₄²⁻ enrichment. Further characterizations of the NH₃-TPD, NO oxidation, and NH₃-SCR performance tests proved that, as a result of the enhanced acidity, the enrichment of SO₄²⁻ on α-Fe₂O₃ displayed a clear correlation with the SCR activity. The results of the present study provide an effective strategy for boosting the catalytic performance of iron oxide in NH₃-SCR via SO₄²⁻ enrichment.

1. Introduction

The selective catalytic reduction of NO with NH₃, which is well known as NH₃-SCR, has been recognized as the most effective technique for removing NOx from stationary sources, such as power plants, industrial boilers, steel mills, and process heaters [1,2]. Currently, the commercial NH₃-SCR catalysts widely used for industrial applications rely heavily on V₂O₅-WO₃(MoO₃)/TiO₂ [3,4]. However, due to the drawbacks of the narrow temperature window and biological toxicity of V₂O₅, there is an urgent need to develop V-free catalysts with a superior NH₃-SCR performance [5,6].

In addition to the high NOx removal efficiency, the development of NH₃-SCR catalysts for practical use must also take into account the influence of SO₂ that is widely at work in the actual operating environments. From this point of view, Fe-based catalysts stand out as a unique candidate due to their outstanding SO₂ resistance [7,8]. For example, Yu et al. [9] conducted a comprehensive study on the SO₂ resistance of the α-Fe₂O₃ catalyst in NH₃-SCR and found that, even at the low reaction temperature of 225 °C, there was negligible activity loss after the long-term test over 216 h. Tang and co-workers [10] reported that, for the catalyst of α-Fe₂O₃ supported on CNTs, the activity showed an obvious increment when SO₂ was introduced. Feng et al. [11] investigated the SO₂ tolerance to Fe₂(SO₄)₃/TiO₂ catalysts and found that, after introducing SO₂ and H₂O for 12 h at 270 °C, the NO conversion was still maintained at 82%.

As a typical transition metal oxide, Fe₂O₃ has good redox properties [12]. However, it usually suffers from weak acidity, which precludes sufficient NH₃ adsorption. As a result, an inferior catalytic performance in NH₃-SCR is exhibited [13]. Interestingly, a variety of previous studies have shown that the sulfation of iron oxide can significantly increase the surface acid site [9,14]. Ma et al. [6] systematically discussed the promotion effect of sulfation in upgrading the catalytic performance of transition metal oxides in NH₃-SCR. In addition, Sun et al. [15] emphasized the important role of the sulfation temperature in promoting the NH₃-SCR performance. Our recent study [16] also revealed that the formation of ferric sulfate species via ABS decomposition can greatly improve the acidic property of Fe₂O₃-based catalysts, making a distinct contribution to the improvement of the activity. As such, it is reasonable to expect that, to promote the catalytic activity of Fe₂O₃-based catalysts in NH₃-SCR, the enrichment of SO₄²⁻ species is an effective method. Unfortunately, up to the present time, studies regarding this issue have rarely been carried out.

It has been widely reported that the building of spatially confined environments can significantly influence the adsorption and diffusion properties of gas molecules [17,18], and this may have some beneficial effects on the immobilization of sulfate species in SO₂ adsorption on NH₃-SCR catalysts. With this purpose in mind, we herein conduct a study on the effects of spatial confinement of mesoporous SBA-15 on the immobilization efficiency of sulfate species derived from ABS decomposition. SBA-15 is such kind of material, with regularly arranged mesopore channels and a uniform pore size, which is highly suitable for investigating the spatially confined effect. Two catalysts, namely Fe₂O₃/SBA-15 and Fe₂O₃/SiO₂, with comparable Fe₂O₃ loading amounts but distinctly different spatial distributions, were fabricated. After ABS loading and high temperature treatment to induce the SO₂ evolution, the effects of spatial confinement on the enriched immobilization of the sulfate species on α-Fe₂O₃ were examined, and a clear correlation with the SCR activity was established.

2. Results and Discussion

2.1. The Encapsulation of Fe₂O₃ in the Mesopores (TEM and N₂ Sorption)

TEM characterization provides direct evidence for the spatial confinement of Fe₂O₃ in the pore channels of SBA-15. As depicted in Figure 1a, the mesopores with a regular arrangement can be clearly observed, demonstrating that the ordered framework is well preserved after Fe₂O₃ loading. On the other hand, based on the dark-field STEM image recorded with a high angular annular dark field (HAADF) detector (Figure 1b), it can be safely concluded that most of the Fe species present as small nanoparticles and, more importantly, they are encapsulated in the pore channels. In contrast to SBA-15, the reference SiO₂ is composed of irregular particles with a size about 50–100 nm (Figure S1). Moreover, the N₂ sorption and pore size distribution results (Figure S2) reveal that it only contains macropores of a size larger than 80 nm, which are probably assembled by large SiO₂ particles. As such, it is reasonable to conclude that no confinement can be constructed using normal SiO₂.

Additional evidence for the location of the Fe₂O₃ nanoparticles in the mesopores can be deduced from the N₂ physisorption results. As can be seen from Figure 2, a steep hysteresis loop characteristic of the ordered mesopores is indicated by the N₂ adsorption/desorption isotherm. Additionally, the desorption branch exhibits a forced closure at p/p⁰ = 0.42, which is not detected in pure SBA-15. According to the literature [19,20], the appearance of this phenomenon is related to the cavitation effect caused by the encapsulation of guest species. This reveals that some Fe₂O₃ particles are present in the mesopores, but the faint cavitation effect indicates they are minuscule and do not act as serious obstructions of the pores [21]. Additionally, the textural parameters of the different samples were collected. Compared with SBA-15 (465 m²/g, 1.16 cm³/g), the specific surface area and pore volume of Fe₂O₃/SBA-15 (391 m²/g, 0.82 cm³/g) displayed an obvious decline. This demonstrates that Fe₂O₃ was successfully introduced into the mesopores by the wet impregnation method.

2.2. Surface Dispersion Properties of Fe₂O₃ in the Fe₂O₃/SBA-15 and Fe₂O₃/SiO₂ Catalysts (XRD and XPS)

Figure 3 shows the XRD results of the two catalysts. In addition to the appearance of a broad reflection at 2θ = 10–30°, attributed to the amorphous structure of the silica [22,23], multiple diffraction peaks at 2θ of 33.2, 35.7, 40.8, 49.7, 54.2, 62.8, and 64.0° were detected, which can be attributed to the characteristic diffractions resulting from α-Fe₂O₃ [24]. Meanwhile, comparing the signal intensities of the two catalysts, no obvious difference was observed. This result indicates that the Fe species were present as the crystalline phase of α-Fe₂O₃ on both carriers and, at the same time, a comparable dispersion degree could be expected.

To gain insight into the chemical state of the Fe species and the relative surface atomic concentrations, XPS characterization was performed. As shown in Figure 4, the Fe 2p profiles of both catalysts contained two peaks with binding energies of around 711.2 and 724.9 eV, which can be attributed to Fe 2p3/2 and Fe 2p1/2, respectively. This is characteristic of Fe species in the +3 oxidation state [25,26]. Moreover, due to the absence of electronic exchange between α-Fe₂O₃ and the inert SiO₂, a negligible shift in the Fe 2p signals for the two catalysts was detected. This indicates the obtainment of a similar chemical state of the Fe species in both catalysts.

Additionally, the dispersion property of α-Fe₂O₃ was estimated by referring to the area ratio of the respective characteristic peaks of Fe and Si. It was found that the surface atomic concentrations of Fe in the two catalysts did not vary greatly. That is, Fe₂O₃/SBA-15 exhibited a surface Fe concentration of 9.7%, as compared to 8.3% in the case of Fe₂O₃/SiO₂. This result demonstrates that the dispersion degree of Fe species is not largely affected by the choice of different SiO₂ supports, which is consistent with the characterization results from the XRD analysis.

2.3. Effects of the Mesopore Confinement on the Capture of SO₂ on Fe₂O₃

The lack of electronic interaction between iron oxide and SiO₂, together with the comparable Fe₂O₃ dispersion, enables the reliable exploration of the spatial confinement effect on the enrichment of SO₄²⁻. Herein, by taking advantage of the highly soluble property of ferric sulfate/ammonium sulfate in water, we used ion chromatography (IC) as an effective tool to obtain quantitative information about the adsorbed sulfate species in the Fe₂O₃ catalysts [9]. To exclude any disturbance from the mesoporous silica affecting the accuracy of the test, pure ABS/SBA-15 and ABS/SiO₂ were dissolved in 20 mL of deionized water, and the measured results showed that 97.2% and 98.3% of the sulfate could be dissolved. As such, we can conclude that SBA-15 and SiO₂ supports have a negligible hindering effect on the accurate measurement of sulfate species.

Along with our previous study [16], ABS was employed in this work to investigate the source of the sulfur species. It is reasonable to assume that, among other factors, the immobilization efficiency depends highly on the amount of SO₂ evolved from the ABS decomposition. Thus, the decomposition behavior of ABS under real working conditions was explored first. As ABS starts to decompose at 300 °C, a high reaction temperature of 400 °C was adopted to trigger the ABS decomposition. After thermal treatment for 0.5 h, the collected samples (named with T- as a prefix for their discrimination) were dissolved in deionized water. The measured concentrations of SO₄²⁻ are shown in

Table 1. The measured concentrations of SO₄²⁻ using ion chromatography for the varied samples.

Samples	Detected SO42− Concentration (mg/L)	Theoretical Maximum Concentration (mg/L)	Extra SO42− Concentration Derived from Fe2O3 Trapping (mg/L)
ABS/SBA-15 (untreated)	38.58	39.71	-
T-ABS/SBA-15 *	18.49	39.71	7.84
T-ABS/Fe2O3/SBA-15	26.33	39.71
ABS/SiO2 (untreated)	39.03	39.71	-
T-ABS/SiO2	9.80	39.71	3.04
T-ABS/Fe2O3/SiO2	12.85	39.71

* The prefix T- attached to the sample name denotes that the catalyst that was treated under the reaction stream at 400 °C for 0.5 h.

. It is clear that about half of the ABS was not decomposed on T-ABS/SBA-15, while only a quarter was found to have decomposed on T-ABS/SiO₂. Previously, our study [27,28] revealed that the location of ABS in a confined environment actually exerts a negative effect on its decomposition, and the decomposition behavior displayed a close pore/cavity-size-dependent pattern. This characteristic appears to make it difficult for the confinement system to produce a significant amount of metal sulfate at a certain temperature.

ABS/Fe₂O₃/SiO₂ and ABS/Fe₂O₃/SBA-15 were treated using the same conditions, and the measured results are listed in Table 1. It makes sense that, with the introduction of α-Fe₂O₃, some of the evolved SO₂ can be captured, and this leads to the detection of further sulfate species. As such, the actual Fe₂O₃ contents in the catalysts were first measured by ICP, and the obtained results show that the Fe contents were close to nominal loading (10.0%) for Fe₂O₃/SiO₂ (11.3%) and Fe₂O₃/SBA-15 (9.6%). Surprisingly, after deducting the SO₄²⁻ from the undecomposed ABS, the mesoporous SiO₂-supported catalyst displayed an enhanced sulfate concentration of 7.84 mg/L, while the corresponding value for the normal SiO₂-supported catalyst was only 3.04 mg/L. This indicates that the spatial confinement of the mesopore channels provides a great benefit in capturing sulfur to form metal sulfate, which is very important for enhancing the SCR reaction activity.

2.4. Effects of Enriched Sulfate on the Acid and Redox Properties of the Catalysts (NH₃-TPD, NO Oxidation)

The effect of sulfate immobilization on the acidity of the catalysts was explored by NH₃-TPD, and the results are shown in Figure 5. Three desorption peaks were observed for the two catalysts. Peak α, centered at 120 °C, can be attributed to the NH₃ desorbed from the physical/weakly adsorbed sites, while the peaks β and γ can be attributed to the NH₃ desorbed from the moderately and strongly acidic sites, respectively [29]. Ma et al. [30] found that sulfation has great influence on the formation of Brønsted acid sites, which contributes to the moderately and strongly acidic sites. It can be seen that the intensities of the peaks β and γ in T-ABS/Fe₂O₃/SBA-15 are much higher than those of the reacted T-ABS/Fe₂O₃/SiO₂, verifying that the enrichment of sulfate species is conducive to promoting the acidic properties of α-Fe₂O₃ catalysts. During the process of NH₃ desorption, an SO₂ signal was also detected when the temperature exceeded 350 °C. It was found that the SO₂ signal intensity released by the T-ABS/Fe₂O₃/SBA-15 catalyst was much greater than that of the T-ABS/Fe₂O₃/SiO₂ catalyst, further supporting the obtainment of enriched SO₄²⁻ via spatial confinement.

In addition, Guo et al. [16] explained that, after ABS deposition and subsequent NH₃-SCR reaction, the reason for the enhanced activity is not only the enhancement of the acidity, but also the contribution of the redox properties. This suggests that the formation of metal sulfate may improve the redox performance of the catalyst. Hence, to characterize the differences between the redox performances of the two catalysts, a NO oxidation experiment was carried out, and the results are shown in Figure S1. A negligible NO oxidation performance was exhibited in the temperature range of 150–400 °C, and both catalysts showed a conversion efficiency of less than 2.5%. The above results clearly show that the formation of a large number of ferric sulfate species is advantageous for the enhancement of the surface acidity of the catalysts, thereby increasing the chemisorption of NH₃.

2.5. Promotion Effect of the SO₄²⁻ Immobilization on the Catalytic Performance of α-Fe₂O₃ in NH₃-SCR

Figure 6 presents the NH₃-SCR performance of the α-Fe₂O₃ catalysts before and after SO₄²⁻ immobilization. For the pure Fe₂O₃/SBA-15 and Fe₂O₃/SiO₂ catalysts, the NO removal efficiency at 350 °C was 55.3% and 49.9%, respectively. Taking into account the slight difference between the degrees of dispersion of the Fe species on the two catalysts, the NO conversion per unit Fe concentration is comparable. The results also suggest that no obvious enhancement of the NH₃-SCR activity can be attained by the spatial confinement environment.

Meanwhile, the effect of the SO₄^2- immobilization on the catalytic activities of Fe₂O₃/SBA-15 and Fe₂O₃/SiO₂ was revealed. It was found that, after high temperature treatment (400 °C), ABS showed no toxic effect on the catalyst [31]. On the contrary, the SCR activity of the T-ABS/Fe₂O₃/SiO₂ catalyst increased from 49.9% to 77.5%, while that of the T-ABS/Fe₂O₃/SBA-15 catalyst increased from 55.3% to 90.6%. Clearly, the promotion effect of SO₄²⁻ on NH₃-SCR performance was confirmed. Moreover, it is noticeable that the efficiency of the catalyst supported by the mesoporous SiO₂ increased by 35.3% at 350 °C, while the efficiency of the catalyst supported by the ordinary SiO₂ only increased by 27.6%. This clearly shows the positive effect of the SO₄²⁻ enrichment on the NH₃-SCR performance of α-Fe₂O₃ catalysts.

Based on the above characterization results, a schematic illustration (Scheme 1) is presented for the enrichment of the SO₄²⁻ immobilization via the spatial confinement effect. On the normal SiO₂, a large amount of ABS was decomposed, but the generated SO₂ was randomly evolved. As a result, after the introduction of α-Fe₂O₃, the majority of the SO₂ evolved due to the ABS decomposition did not directly interact with α-Fe₂O₃, and only a tiny quantity of SO₂ was captured to form iron sulfate and remained on the catalyst surface. In contrast, for the mesoporous SBA-15, only a small portion of ABS was disintegrated, but the evolved SO₂ was confined by the mesopore channels, which are particularly beneficial for the interaction of SO₂ with FeOx species. Thus, after α-Fe₂O₃ was introduced into the mesopores, even if only a smaller quantity of ABS was decomposed, the majority of SO₂ was captured by the Fe species. As a result, more ferric sulfate was immobilized on the catalyst surface. The results of the present study convincingly illustrate the positive effect of spatial confinement on the enriched immobilization of SO₄²⁻ on α-Fe₂O₃ catalysts for robust denitration.

3. Experimental Section

3.1. Catalysts Preparation

SiO₂ (AR) was purchased from the supplier. SBA-15 with periodic mesopores was prepared according to literature [28,32]. Typically, 4.0 g of P123 (polyethylene oxide-polypropylene oxide-polyethylene oxide triblock copolymer) was dissolved in 30 mL of deionized water, and 20 mL of 2.0 M HCl was added. The solution was stirred until it was clear. Then, under strong stirring, 9.0 g TEOS (tetraethyl orthosilicate) was slowly added dropwise into the mixture, which continued to be stirred in an oil bath at 40 °C for 24 h. The milky slurry was transferred into a stainless-steel reaction kettle with a Teflon lining and treated using a typical hydrothermal process at 100 °C for 24 h. After filtration and washing with deionized water and ethanol several times, the original powder of SBA-15 was obtained. To remove the surfactant, the as-prepared sample was calcined at 550 °C for 5 h.

The loading of the active component, Fe₂O₃, was accomplished by wet impregnation. The calculated amount of Fe(NO₃)₃·9H₂O (10 wt.%) was dissolved in 30 mL of deionized water. With the addition of the designated support (SiO₂ and SBA-15), the mixture was stirred for 2 h. After evaporating the excess water at 80 ^oC in an oil bath, the obtained samples were dried overnight in an oven at 100 °C, followed by calcination at 450 °C for 4 h in an air atmosphere. For the sake of simplicity, the prepared catalysts were labeled as Fe₂O₃/SiO₂ and Fe₂O₃/SBA-15.

Ammonium bisulfate (ABS) was employed as the source of the sulfur species to explore the effect of the support structure on the capture efficiency. The introduction of ABS onto SiO₂, SBA-15, Fe₂O₃/SiO₂ and Fe₂O₃/SBA-15 was also realized by the wet impregnation method. Typically, the calculated amount of ABS (5 wt.%) and support were mixed and dispersed in 30 mL deionized water. Then, the suspension was stirred for 2 h. The excess liquid was evaporated at 80 °C, followed by drying overnight at 100 ^oC. For convenience, the samples were labeled as ABS/SiO₂, ABS/SBA-15, ABS/Fe₂O₃/SiO₂, and ABS/Fe₂O₃/SBA-15.

3.2. Catalyst Characterization

The N₂ physisorption was measured at −196 °C using a Micromeritics ASAP 2460 instrument. The catalysts were degassed for 360 min at 100 °C in the degas port of the adsorption analyzer. The mesoporous distribution was tested using the BJH method.

The powder X-ray diffraction (XRD) patterns of the samples were measured using a Philips X’pert X-ray diffractometer operating with Ni-filtered Cu Kα1 radiation (0.15408 nm) at the 2θ range from 10–80°. The X-ray tube operated at 40 kV and 40 mA.

Transmission electron microscopy (TEM) images were taken on a Tecnai G2 F30 instrument at an accelerating voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) measurements were performed on a PHI 5000 Versa probe system with monochromatic Al Kα radiation (1486.6 eV, 15 kW). All binding energies were calibrated with chance C1 (284.6 eV).

Ion chromatography (ICS-900) was used to detect the immobilized SO₄²⁻ on the samples. Typically, 20 mg of the sample was dispersed in deionized water (20 mL). The aged solution was filtered with a 0.22 μm filter membrane before the IC measurement.

The NH₃ temperature-programmed desorption (NH₃-TPD) measurements were carried out to evaluate the acid properties of the catalysts. The device was equipped with an FTIR spectrometer to monitor the outlet gas signals, and the time interval for the spectral acquisition was set to 30 s. Before the tests were carried out, the sample was purged with O₂ at 150 °C for 30 min and cooled to room temperature. During the adsorption process, 40 mL/min of NH₃ and 60 mL/min of equilibrium gas Ar were introduced at room temperature until saturation was achieved. The sample was then purged with Ar for 2 h to remove the residual gas. The tests were performed by increasing the temperature from 30 °C to 600 °C at a ramp rate of 10 °C/min and an Ar flow rate of 100 mL/min.

The redox properties of the catalysts were evaluated by NO oxidation experiments. Before the tests, the samples were purged with O₂ at 150 °C for 30 min and cooled to room temperature. Next, the NO oxidation was performed using the same experimental setup as the NH₃-SCR performance.

3.3. Activity Measurements

The NH₃-SCR performance of the catalysts was assessed using a fixed-bed reactor with a reaction gas composition of 500 ppm NO, 500 ppm NH₃, 5 vol.% O₂, and Ar as the equilibrium gas, and the total gas flow was maintained at 100 mL/min. First, a certain quantity (100 mg) of the sample was sieved to a 60–80 mesh size. Then, the catalyst was loaded into a glass tube reactor. It was purged at 150 °C by O₂ for 0.5 h and then cooled to room temperature, followed by cutting in the reaction gas adsorption until saturation. The reaction was carried out at a space velocity of 60,000 mL·g⁻¹·h⁻¹. The concentration of NO was detected using a nitric oxide sensor (Citicel). Then, the catalytic performance data were collected after maintaining stability for 30 min at each temperature point. The NO conversion was obtained using the following equation:

$$\mathrm{NO} \mathrm{Conversion} \left(\%\right)=\frac{{\left[\mathrm{NO}\right]}_{\mathrm{in}}-{\left[\mathrm{NO}\right]}_{\mathrm{out}}}{{\left[\mathrm{NO}\right]}_{\mathrm{in}}}\times 100\%$$

4. Conclusions

In the present study, a novel strategy for promoting the immobilization of sulfate species on α-Fe₂O₃ via spatial confinement for an improved NH₃-SCR performance was reported. We intentionally loaded ABS onto two designated catalysts (Fe₂O₃/SBA-15 and Fe₂O₃/SiO₂) with the same Fe₂O₃ loading capacity but distinctly different spatial distributions, and the effects of mesopore confinement on the immobilization efficiency of SO₂ on α-Fe₂O₃ was explored. In comparison with normal SiO₂, it was found the employment of the mesoporous silica SBA-15 did not show any advantage in terms of the ABS decomposition, but with the assistance of the spatial confinement environment, it exhibited the great benefit of capturing the evolved SO₂ from ABS decomposition to form metal sulfate. The enriched SO₄²⁻ immobilization on α-Fe₂O₃ greatly promoted the surface acidity of α-Fe₂O₃, and more NH₃ species were available to participate in the NH₃-SCR reaction. Consequently, the denitration efficiency was greatly enhanced.